In computer software, semantic versioning (or something like it) is something of a standard for how to version software releases. The Major.Minor.Patch semantics make clear how big of a change has occurred between the present and an earlier release. Is there something similar for versionining a dataset? Unlike with software, any change to a dataset might be a "major" change, so I'm not sure that the logic translates directly. Is there any commonly used specification for versioning a dataset?
Asked
Active
Viewed 472 times
12
Thomas
- 1,114
- 6
- 14
-
I wonder if the dat folks have thought about this http://dat-data.com/ max ogden & co. - I don't know, but i imagine they'd have some insight here. – sckott Aug 17 '15 at 18:59
-
BTW, https://frictionlessdata.io/specs/patterns/#data-package-version – Stanislav Kralin Apr 13 '18 at 07:51
2 Answers
8
Franck Dernoncourt
- 7,780
- 9
- 39
- 86
-
1The link above is a 404 but the image can now be found at http://guides.dataverse.org/en/4.8.6/user/dataset-management.html#dataset-versions – Philip Durbin Apr 13 '18 at 00:49
3
The Digital Curation Centre guide on how to cite datasets and link to publications gives a good summary to the various approaches to dataset versioning from the viewpoint of repository managers, and links to further information.
-
2Please provide a précis of the content you linked in this post, as links may change and web resources become inaccessible over time. – Aug 17 '15 at 20:32